back to article Oracle: You wanted SQL on Hadoop? How about SQL on Hadoop and SQL ON NOSQL?

Oracle wants all technology to speak the same language as the one found in its eponymous database. Unfortunately for Oracle, the world doesn't work like that, so on Tuesday the company announced new software that lets Oracle aficionados use Oracle SQL commands to probe non-SQL systems, such as Hadoop and NoSQL databases. The …

  1. cs94njw

    This rings a bell. This smacks of Microsoft's "Embrace & Extend" tactics, which introduced incompatibilities and dragged customers away from the original technology.

    *shrug* It's certainly tempting to use a proper SQL language...

    1. Swarthy

      It is tempting to use a proper SQL language

      ...When do you think Oracle will implement one?

  2. Joerg

    NoSQL and other crap..all scams and frauds...

    A lot of marketing babbling around NoSQL and other "new standards" ... Fact is that these are far from better than SQL unlike the marketing of all multinationals behind it want people to believe.

  3. Otto is a bear.


    Well not really a surprise, but then Big Data is still not as evolved as the RDBMS market, think around the Mid 80s for tool support on RDBMS and front end wise that's where most BD/BoSQL suppliers are. There are some neat 3rd party tools out there, though.

    Trouble is BD & NoSQL are the Buzz Words du jour, so the answer must be Hadoop, or Mongo, or .....

    It's nice to have a familiar tool, but actually SQL doesn't really lend itself to big data questions, anymore than big data lends its self to answering relational questions.

    Still keeps me employed so I'm not complaining.

    1. Anonymous Coward
      Anonymous Coward

      Re: Hmmmm

      I think the idea here is that you can have your cake and eat it. The data is held in HDFS and described in HCatalog so you can query with any or all of MapReduce, Pig, Spark, Impala or Bigdata SQL. You can hold the data in any format you like as long as you can define a SerDe ( in HCatalog so supports stuff like avro and parquet with hadoop built-in SerDe.

      So for those analyses which are best implemented in procedural code, then use something from the Hadoop family, if it's best done in a declarative SQL style, then thats fine too - use Bigdata SQL. No need to move the data or reformat.

  4. Tom 64

    Sounds like a 'look, me too!' endeavour.

    Apache Hive actually already allows you to do most SQL operations on data in HDFS, and is enough for our analysis needs. Being able to join on data not in HDFS does however sound interesting.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like